CN2854594Y - Cutter working condition monitoring device based on voice identification technology - Google Patents
Cutter working condition monitoring device based on voice identification technology Download PDFInfo
- Publication number
- CN2854594Y CN2854594Y CN 200520126693 CN200520126693U CN2854594Y CN 2854594 Y CN2854594 Y CN 2854594Y CN 200520126693 CN200520126693 CN 200520126693 CN 200520126693 U CN200520126693 U CN 200520126693U CN 2854594 Y CN2854594 Y CN 2854594Y
- Authority
- CN
- China
- Prior art keywords
- signal
- speech recognition
- cutter
- monitoring device
- device based
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Expired - Fee Related
Links
Landscapes
- Investigating Or Analyzing Materials By The Use Of Ultrasonic Waves (AREA)
Abstract
The utility model discloses a knife working condition monitoring device based on speech recognition technology. The device has solved the problem that the knife monitoring in current manufacture industry is not able to meet the actual demand. In addition, the utility model is characterized by simple structure and convenient use, and available of using a plurality of knives, etc. The structure of the device is as follows: the device comprises a collection device of the knife cutting sound signal, and this device collects various knife's cutting sound feature signals and the real-time sound signal of the inspected knife; the collection device is connected with the pre-play and amplification control device, converts the sound signal into analog/ digital, amplifies and the eliminates the noise, and sends the generated digital signal into the speech recognition device. The device inputs the feature knife cutting signal into the memory device as the comparison benchmark for the real-time signal; the real-time cutting signal of the detected knife will be matched with the comparison benchmark signal, and then the comparison result will be input into the output device.
Description
Technical field
The utility model relates to a kind of monitoring device of machine tool, relates in particular to a kind of cutter operating mode monitoring device based on speech recognition technology.
Background technology
Manufacturing industry is the basic industry that promotes social development and technical progress, and it is affecting the every aspect of people's lives, is concerning national economy, and cutter is the critical component in the manufacturing industry, and it is to manufacturing development decisive role.In the process of metal cutting processing, the mill of cutter, damaged as fail to find timely can cause the interruption of working angles, the quality that causes workpiece descend or scrap, lathe damages, thereby may cause enormous economic loss.Therefore, in the machining process, quantitatively, periodically grasp cutting tool state, monitoring and diagnosis tool wear, tipping equivalent damage fault, for prolonging machine tool failure-free operation, improving the quality of products has great importance.
In recent decades, aspect tool monitoring, obtained bigger progress both at home and abroad, all reached certain level on the technical degree of depth and the range, but also there is not a kind of method can be applicable to the monitoring of various cutters under the different processing conditionss, just the whole bag of tricks restricted application also reaches the requirement of robotization control far away.By the check and the analysis of actual engineering, find that the common defective of existing tool monitoring method is: the randomness of the diversity of cut and cutter, the polytrope of cutting parameter, cutter blunt and the contradiction of monitor signal monotonicity are still unresolved.Performances such as the fast-response of cutter operating mode monitoring method, reliability, robustness can't satisfy actual requirement fully.
Summary of the invention
The utility model is exactly in order to address the above problem, and provides a kind of and has simple in structurely, easy to use, can use the cutter operating mode monitoring device based on speech recognition technology of advantage such as multiple cutter.
For achieving the above object, the utility model has adopted following technical scheme: a kind of cutter operating mode monitoring device based on speech recognition technology, it comprises the harvester of Tool in Cutting acoustical signal, and this device is gathered the diagnostic acoustic signal of various Tool in Cutting and the real-time acoustical signal of institute's detection cutter; This harvester is connected with putting in advance with gain control, acoustical signal is carried out amplifying after mould/number conversion, removed and make an uproar, the digital signal that produces is sent into speech recognition equipment, and this device is sent feature Tool in Cutting acoustical signal into memory storage, as the benchmark of real-time acoustical signal; The real-time cutting acoustical signal of detected cutter and this benchmark signal mate, and comparative result is exported by input-output unit.
Described cutter acoustical signal harvester comprises the condenser type sound pressure sensor that has filter bowl that is installed in the lathe tools relevant position, and it gathers remote sound in 20 °～30 ° scopes in dead ahead, and its back side remolding sensitivity is 1: 31.
Described sound pressure sensor is connected with a transmitter.
Described putting in advance with gain control comprises by mould/number conversion circuit, amplifying circuit and filtering circuit.
Described speech recognition equipment is that model is the RSC-4x microprocessor, and it is connected with memory storage, and is connected with output unit by the I/O bus.
Described memory storage is included as the outer RAM of sheet, and it is connected with data bus, address bus, the control bus of microprocessor and the RAM and the ROM of microprocessor internal.
Described input-output unit comprises display device, keyboard and external interface.
Described display device comprises display and acoustic-optic alarm, speech recognition equipment with the matching result of acoustical signal send into show dress after, by the display display result, after the breakage of cutter or sharp wear are identified, acoustic-optic alarm will send alerting signal; Keyboard is used for to speech recognition equipment input configuration information; External interface is the RS232/485 communicating circuit, and can be connected the adaptive control that realizes cut in the mode of serial communication with the control system of numerically-controlled machine by this interface.
The condenser type sound pressure sensor that monitoring device of the present utility model adopted is easy to install, and the position of installing needn't basic cutter or workpiece just can collect signal effectively, thereby be difficult for process is exerted an influence, also can not hinder the loading and unloading of tool changing and workpiece simultaneously; The acoustical signal of being gathered is not vulnerable to smear metal, the influence of heat eliminating medium; That the CPU of voice recognition unit adopts is RSC-4x dedicated voice identification DSP, can carry out quick computing, thereby can reach the purpose of real-time monitoring; The recognizer of monitoring device has adopted intellectual technologies such as artificial neural network, thereby has satisfied the intelligent demand of system.
In the training study process of tool wear diagnostic acoustic signal, can be various machined parameters, the diagnostic acoustic signal of various cutters extracts in advance, be used for training and identification, when monitoring, go for the monitoring of various machined parameters and various cutter situations, thereby remedied in the existing tool monitoring method the defective that can't effectively monitor cutter and machined parameters polytrope; Monitoring system is with Hidden Markov Model (HMM), dynamic regular model is as speech recognition modeling, these models choose the reliability that has guaranteed monitoring system, make cutter in real time the discrimination of monitoring satisfied actual requirement, and suitable speech recognition modeling has also guaranteed the requirement of quick response.
This monitoring device is based on speech recognition technology, with the acoustical signal that produces in the Tool in Cutting process as monitoring variable.Gather the diagnostic acoustic signal of tool wear by the condenser type sound pressure sensor, diagnostic acoustic signal was said that training study was stored in the storer of supervising device afterwards.In real-time observation process, when the characteristic signal of tool wear signal that monitors and training was complementary, voice recognition unit just can identify, and the result of identification is sent to display, thereby can know which kind of the degree of wear cutter reached.
The beneficial effects of the utility model are: 1) speech recognition technology is used for the mechanical processing process measurement and control area, make speech recognition technology obtain wider range of application, and acoustical signal is a kind of extraordinary information source as a kind of information carrier of cutter process state.Sound from identification people's sound to identification process operating mode is for the cutter monitoring of working condition provides a kind of new effective ways.2) the method monitoring cutter machining state that combines with Principles of Acoustics and signal Processing shows, has corresponding relation between Tool in Cutting sound spectrum feature and its wearing and tearing, the damaged situation.Monitor the method for the state of wear of cutter with sound as monitor signal, have the characteristics of accuracy and stability.3) based on speech recognition special DSP platform, adopt HMM and ANN recognizer, the discrimination of acoustical signature signal after training of tool failure fracture can be reached more than 97%.
Description of drawings
Fig. 1 is a structured flowchart of the present utility model;
Fig. 2 is an electrical schematic diagram of the present utility model.
Embodiment
Embodiment:
As shown in Figure 1, this monitoring device mainly contains following unit composition:
1, the harvester of Tool in Cutting acoustical signal
It has adopted the condenser type sound pressure sensor that has filter bowl, and has highly directive, has more sharp-pointed directivity than sharp cardiod microphone, is fit to pick up remote sound in 20 °～30 ° scopes in microphone dead ahead.Just, the ratio of back side sensitivity is 1: 31, this sensor has guaranteed farthest to have avoided other direction to pass the acoustical signal useless to monitoring system of coming simultaneously again to effectively the picking up of Tool in Cutting acoustical signal; Sensor causes actual effect for the splash that prevents heat eliminating medium, on the basis that guarantees effective pick-up of acoustic signals, sensor is installed a filter bowl, so just can avoid the splash of heat eliminating medium effectively.
2, put in advance and gain control
This device comprises that removing of acoustical signal made an uproar, amplify in advance, automatic gain control and mould/number conversion.Therefore the acoustical signal of sensor acquisition is a simulating signal, at first acoustical signal is converted to the accessible digital signal of system by mould/number conversion chip, amplifies in advance then, automatic gain control, at last by wave filter place to go noise, extracts characteristic parameter.Comprise in a large number to the useless acoustical signal signal of monitoring cutting-tool wear state in the acoustical signal that sensor picks up, these unvanted sound signals all can be defined as noise.If noise signal is too big, signal to noise ratio (S/N ratio) is reduced, the monitor signal of cutting-tool wear state may be flooded by these noise signals, thereby is unfavorable for the training study of diagnostic acoustic signal and the identification of cutting acoustical signal.Thereby the acoustical signal of gathering must remove and make an uproar, and essence is exactly in order to improve the signal to noise ratio (S/N ratio) of signal, the discrimination of raising system.The characteristic parameter that extracts is sent to microprocessor and trains and discern.
3, memory storage
Before system monitors the cutter operating mode, gather earlier the diagnostic acoustic signal of tool wear, these acoustical signals are carried out after the pre-service, be sent to voice recognition unit and train and learn, after training successfully, these diagnostic acoustic signals are sent in the storer to be preserved.In order to guarantee memory space and recognition speed, storer has adopted outer RAM of sheet and ram in slice, by selecting suitable searching algorithm, just can guarantee the requirement of monitoring in real time.
4, speech recognition equipment
Voice recognition unit adopts the RSC-4x microprocessor, it is a kind of novel speech recognition and analog input output processor, utilize chip integration to become the 8-bit microprocessor of the audio processing modules of ADC, pre-amplifying circuit, RAM, ROM and integrated optimization, to source pointer and the object pointer that all instructions have complete symmetry, avoided the limitation of A, B and DPTR general-purpose register.Utilize RSC-4X to have high performance and go up characteristic and special-purpose DSP engine, for developing application and tailor-made algorithm provide integrated signal platform efficiently.Adopt the advanced person's of RSC-4x support speech recognition algorithm, move the new HMM algorithm and the neural network algorithm of enhancing, significantly improved the degree of accuracy of in various noises, discerning, strengthen the recognition capability of nonspecific sound, specific sound and continuous sound under close voice and noise circumstance, and be applied in the identification of the characteristic event signal in the continuous sound wave signal.
Utilize these characteristics, when sound that monitors number and diagnostic acoustic signal in the storer are complementary, then export recognition result, the result of identification is sent to display unit, thereby can learn which kind of the degree of wear cutter reached.
5, input-output unit
Input-output unit comprises display device, keyboard and external interface.Display unit comprises display and acoustooptic alarm system.Speech recognition equipment is sent to display device with the matching result of acoustical signal, in this device, is people output result to understand with the conversion of signals that sends; After the breakage of cutter or sharp wear are identified, in this device, not only be converted into displayable result, and send alerting signal, can point out people to change cutter as soon as possible like this, be affected in order to avoid make to produce.Keyboard is used for importing configuration information, as monitoring threshold is set, edit-modify system program etc.External interface is used for expanding monitoring system functional, can carry out serial communication by this interface, carries out long-range setting and Centralized Monitoring, can be connected the adaptive control that realizes cutting process with the control system of numerically-controlled machine.
The condenser type sound pressure sensor that this monitoring system adopted is easy to install, and the position of installing needn't basic cutter or workpiece just can collect signal effectively, thereby be difficult for process is exerted an influence, the while also can not hinder the loading and unloading of tool changing and workpiece; The acoustical signal of being gathered is not vulnerable to smear metal, the influence of heat eliminating medium; That the CPU of voice recognition unit adopts is RSC-4x dedicated voice identification DSP, can carry out quick computing, thereby can reach the purpose of real-time monitoring; The recognizer of monitoring system has adopted intellectual technologies such as artificial neural network, thereby has satisfied the intelligent demand of system.
In the training study process of tool wear diagnostic acoustic signal, can be various machined parameters, the diagnostic acoustic signal of various cutters extracts in advance, be used for training and identification, when monitoring, go for the monitoring of various machined parameters and various cutter situations, thereby remedied in the existing tool monitoring method the defective that can't effectively monitor cutter and machined parameters polytrope; Monitoring system is with Hidden Markov Model (HMM), dynamic regular model is as speech recognition modeling, these models choose the reliability that has guaranteed monitoring system, make cutter in real time the discrimination of monitoring satisfied actual requirement, and suitable speech recognition modeling has also guaranteed the requirement of quick response.
This monitoring system is based on speech recognition technology, with the acoustical signal that produces in the Tool in Cutting process as monitoring variable.Gather the diagnostic acoustic signal of tool wear by the condenser type sound pressure sensor, diagnostic acoustic signal was said that training study was stored in the storer of supervisory system afterwards.In real-time observation process, when the characteristic signal of tool wear signal that monitors and training was complementary, voice recognition unit just can identify, and the result of identification is sent to display, thereby can know which kind of the degree of wear cutter reached.
Claims (8)
1, a kind of cutter operating mode monitoring device based on speech recognition technology, it is characterized in that: it comprises the harvester of Tool in Cutting acoustical signal, this device is gathered the diagnostic acoustic signal of various Tool in Cutting and the real-time acoustical signal of institute's detection cutter; This harvester is connected with putting in advance with gain control, acoustical signal is carried out amplifying after mould/number conversion, removed and make an uproar, the digital signal that produces is sent into speech recognition equipment, and this device is sent feature Tool in Cutting acoustical signal into memory storage, as the benchmark of real-time acoustical signal; The real-time cutting acoustical signal of detected cutter and this benchmark signal mate, and comparative result is exported by input-output unit.
2, the cutter operating mode monitoring device based on speech recognition technology according to claim 1, it is characterized in that: described cutter acoustical signal harvester comprises the condenser type sound pressure sensor that has filter bowl that is installed in the lathe tools relevant position, it gathers remote sound in 20 °～30 ° scopes in dead ahead, and its back side remolding sensitivity is 1: 31.
3, the cutter operating mode monitoring device based on speech recognition technology according to claim 2, it is characterized in that: described sound pressure sensor is connected with a transmitter.
4, the cutter operating mode monitoring device based on speech recognition technology according to claim 1 is characterized in that: described putting in advance with gain control comprises by mould/number conversion circuit, amplifying circuit and filtering circuit.
5, the cutter operating mode monitoring device based on speech recognition technology according to claim 1 is characterized in that: described speech recognition equipment is that model is the RSC-4x microprocessor, and it is connected with memory storage, and is connected with output unit by the I/O bus.
6, the cutter operating mode monitoring device based on speech recognition technology according to claim 1, it is characterized in that: described memory storage is included as the outer RAM of sheet, and it is connected with data bus, address bus, the control bus of microprocessor and the RAM and the ROM of microprocessor internal.
7, the cutter operating mode monitoring device based on speech recognition technology according to claim 1, it is characterized in that: described input-output unit comprises display device, keyboard and external interface.
8, the cutter operating mode monitoring device based on speech recognition technology according to claim 7, it is characterized in that: described display device comprises display and acoustic-optic alarm, after speech recognition equipment is sent the matching result of acoustical signal into the demonstration dress, by the display display result, after the breakage of cutter or sharp wear are identified, acoustic-optic alarm will send alerting signal; Keyboard is used for to speech recognition equipment input configuration information; External interface is the RS232/485 communicating circuit, and can be connected the adaptive control that realizes cut in the mode of serial communication with the control system of numerically-controlled machine by this interface.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN 200520126693 CN2854594Y (en) | 2005-12-31 | 2005-12-31 | Cutter working condition monitoring device based on voice identification technology |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN 200520126693 CN2854594Y (en) | 2005-12-31 | 2005-12-31 | Cutter working condition monitoring device based on voice identification technology |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN2854594Y true CN2854594Y (en) | 2007-01-03 |
Family
ID=37581231
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN 200520126693 Expired - Fee Related CN2854594Y (en) | 2005-12-31 | 2005-12-31 | Cutter working condition monitoring device based on voice identification technology |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN2854594Y (en) |
Cited By (9)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101614583B (en) * | 2009-04-30 | 2011-06-22 | 上海电力学院 | Plugging material detection early warning device and method based on voice recognition |
| CN105190304A (en) * | 2013-04-19 | 2015-12-23 | 卡特彼勒公司 | Erosion monitoring system for ground engaging tool |
| CN103852525B (en) * | 2012-11-29 | 2016-09-21 | 沈阳工业大学 | Acoustic emission signal recognition methods based on AR-HMM |
| CN106042004A (en) * | 2016-05-30 | 2016-10-26 | 京东方科技集团股份有限公司 | Abrasion alarming device and method |
| CN106371407A (en) * | 2015-07-20 | 2017-02-01 | 陈益民 | Statistic method for machining working state and machining robot |
| CN107756139A (en) * | 2016-08-18 | 2018-03-06 | 王智中 | Tool state detection system and method |
| CN109084887A (en) * | 2018-06-01 | 2018-12-25 | 上海大学 | A kind of pruner blade mechanisms noise simulation method |
| CN110297042A (en) * | 2019-06-28 | 2019-10-01 | 拓卡奔马机电科技有限公司 | A kind of cutting knife-breaking detecting method, device and equipment based on voice recognition |
| CN114184684A (en) * | 2021-09-26 | 2022-03-15 | 无锡朝信科技信息系统有限公司 | Online detection method and system for cutting state of cutter |
-
2005
- 2005-12-31 CN CN 200520126693 patent/CN2854594Y/en not_active Expired - Fee Related
Cited By (11)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101614583B (en) * | 2009-04-30 | 2011-06-22 | 上海电力学院 | Plugging material detection early warning device and method based on voice recognition |
| CN103852525B (en) * | 2012-11-29 | 2016-09-21 | 沈阳工业大学 | Acoustic emission signal recognition methods based on AR-HMM |
| CN105190304A (en) * | 2013-04-19 | 2015-12-23 | 卡特彼勒公司 | Erosion monitoring system for ground engaging tool |
| CN106371407A (en) * | 2015-07-20 | 2017-02-01 | 陈益民 | Statistic method for machining working state and machining robot |
| CN106042004A (en) * | 2016-05-30 | 2016-10-26 | 京东方科技集团股份有限公司 | Abrasion alarming device and method |
| CN106042004B (en) * | 2016-05-30 | 2018-09-11 | 京东方科技集团股份有限公司 | A kind of wear abrasion alarming device and method |
| CN107756139A (en) * | 2016-08-18 | 2018-03-06 | 王智中 | Tool state detection system and method |
| TWI637169B (en) * | 2016-08-18 | 2018-10-01 | 王智中 | Tool status detection system and method thereof |
| CN109084887A (en) * | 2018-06-01 | 2018-12-25 | 上海大学 | A kind of pruner blade mechanisms noise simulation method |
| CN110297042A (en) * | 2019-06-28 | 2019-10-01 | 拓卡奔马机电科技有限公司 | A kind of cutting knife-breaking detecting method, device and equipment based on voice recognition |
| CN114184684A (en) * | 2021-09-26 | 2022-03-15 | 无锡朝信科技信息系统有限公司 | Online detection method and system for cutting state of cutter |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN2854594Y (en) | Cutter working condition monitoring device based on voice identification technology | |
| CN109396953B (en) | Machine tool working state intelligent identification system based on signal fusion | |
| CN103558029B (en) | A kind of engine abnormal noise on-line fault diagnosis system and diagnostic method | |
| CN110160765B (en) | Impact characteristic identification method and system based on sound or vibration signal | |
| CN106112697B (en) | A kind of milling parameter automatic alarm threshold setting method based on 3 σ criterion | |
| CN104723171A (en) | Cutter wear monitoring method based on current and acoustic emission compound signals | |
| Li et al. | Multi-scale statistical process monitoring in machining | |
| CN101566541A (en) | System for evaluating tensile damage of in-service 16Mn steel force-bearing part by adopting catastrophic model | |
| CN110031552B (en) | Structural health monitoring damage characteristic value calculation method | |
| CN110561195A (en) | Method for monitoring flutter in machining process | |
| CN104101652A (en) | Audio signal based wind power blade damage monitoring method and system | |
| CN112207631B (en) | Method for generating tool detection model, method, system, device and medium for detecting tool detection model | |
| CN104708497A (en) | Tool wear monitoring system based on current and sound emission composite signals | |
| CN204525045U (en) | A kind of Tool Wear Monitoring system based on electric current and sound emission composite signal | |
| CN114137410A (en) | Hydraulic mechanism circuit breaker fault identification method based on voiceprint detection technology | |
| CN114850969A (en) | Cutter failure monitoring method based on vibration signals | |
| CN108956783B (en) | HDP-HSMM-based grinding sound grinding wheel passivation state detection method | |
| CN107037133A (en) | A kind of combination sound emission and the machining state monitoring method of current detecting | |
| CN108205016A (en) | The industrial computer system of sound emission safety of tower crane detection | |
| CN113053412A (en) | Sound-based transformer fault identification method | |
| CN106548771A (en) | For the method that speech recognition system eliminates burst noise | |
| CN115376302A (en) | Fan blade fault early warning method, system, equipment and medium | |
| CN201322975Y (en) | Working roll surface damage online monitoring device in continuous mill | |
| CN218973624U (en) | Noise monitoring and positioning system | |
| CN207703792U (en) | Sound emission safety of tower crane detecting system |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C14 | Grant of patent or utility model | ||
| GR01 | Patent grant | ||
| C17 | Cessation of patent right | ||
| CF01 | Termination of patent right due to non-payment of annual fee |
Granted publication date: 20070103 Termination date: 20101231 |